Slaying the gerrymander
(Welcome, New York Times readers!) Thanks to commenters on this topic. Your feedback has shaped my thinking on this subject. I recall bein...
Senate: 50 Dem | 50 Rep (range: 47-52)
Control: D+0.3% from toss-up
Generic polling: D+0.8%
Control: D+0.8%
Harris: 274 EV (D+0.3% from toss-up)
Moneyball states: President AK AZ NC
Click any tracker for analytics and data
(cross-posted with my new Substack)
Lots of pixels have been spilled on a legal theory once considered fringe, the Independent State Legislature doctrine. This theory threatens to wreak havoc with centuries of election law. Two upcoming Supreme Court redistricting cases cite this doctrine. Both are brought by Republican-controlled legislatures, so you’d think it would be of net benefit to their party. I did the math. Like the web ads say, the answer might surprise you.
Until recently, the theory was considered radical. It is based on Article I and II of the Constitution, which assign state-level power to regulate federal elections to legislatures. In two cases, Moore v. Harper and Costello v. Carter, lawyers representing Republican legislators question how much of a role the word “legislature” leaves for state courts.
A favorable ruling would go against precedents going back to George Washington’s first term of office. Ironically, the North Carolina General Assembly itself, represented in Moore v. Harper, passed a law two decades ago explicitly handing authority over redistricting to the state Supreme Court they now oppose. But recent Supreme Court rulings – on abortion, on religious expression, and on election law – make clear that the Court is unafraid to break from the past when there’s power or policy at stake.
If the Supreme Court does rule in favor of the theory, voters across the nation will take a major hit in the form of fewer competitive seats. Legislators generally draw safer districts than courts or independent commissions. I have analyzed ten states whose redistricting will potentially be affected by the ISL doctrine. In these states, up to 25 competitive Congressional districts would be replaced by single party-drawn plans. That’s over half of the competitive districts drawn this year! (Contrary to what is being said out there, there’s actually a fair lot of competition in the new maps. That’s a topic for a different day.)
The Independent State Legislature theory would disrupt the partisan balance that has emerged in many states. To determine this, I estimated what a party-blind redistricting process would produce in the ten states* that would be currently affected by the theory. The Princeton Gerrymandering Project has performed computer simulations** and used fairness metrics to identify neutral ranges of outcomes. Only two states, both under single-party control, have outcomes outside the range, Florida (R) and Maryland (D). Left unchecked, single-party control would let the other eight states join them.
However, Congressional power is determined not by single states, but by their combined total representation. So let’s add it all up.
In Moore, overturning court oversight in North Carolina might net three Congressional seats for Republicans. But these seats would potentially be offset by two other states: New York and Maryland. In New York, a court-appointed special master redrew a Congressional map drawn by Democrats that is expected to elect between one and three more Republicans than the legislature’s map. In Maryland, a Democratic seat was made competitive. Applying a win in Moore to all three states would lead to nearly no change in power. Net change: zero, plus or minus one.
A more radical solution is proposed by Pennsylvania legislators in Costello, who suggest that in case of an impasse between the legislature and governor, state courts should be constrained by a federal law that gives a fallback plan. Representatives would be elected under the old lines, or if the number of representatives changes due to reapportionment, by statewide at-large election. In 2020, Pennsylvania was won by Joe Biden, and lost a seat with the new Census. At-large elections would likely have given Democrats an additional nine seats. Whichever way Pennsylvania votes in 2024, the stakes would be increased enormously. Net change: up to nine seats, but which way?
Although it’s not mentioned in the cases before the Supreme Court, there are additional interpretations of the Independent State Legislature theory. Which party do these interpretations help?
First, the theory might eliminate citizen redistricting commissions. Currently, citizens can bring initiatives to the ballot in about 25 states. The establishment of independent redistricting commissions by this route has been a success story for reform. Voters have used this mechanism in Arizona, California, and Michigan. For reformers of both parties and no party, it would be a bitter pill if power reverted to the legislature. In Arizona, the enacted map gets an A grade from the Princeton Gerrymandering Project, with two Democratic seats, four Republican seats, and three competitive seats. An all-Republican legislative process might turn those three competitive seats red. But in California, with its 52 total Congressional districts, the Democratic legislature and governor would be able to draw a map that elects four or five additional Democrats. So overturning independent commissions in these two states might net Democrats an additional one or two seats. (It is worth noting that in Michigan, benefits would go to whichever party wins this fall’s legislative and governor’s elections. This is one case in which Republicans stand to gain, since Democrats seem unlikely to overturn the work of an independent commission.)
In perhaps the most aggressive interpretation of the independent state legislature theory, the Court could say that a governor cannot veto a legislature’s actions. This year, two Congressional redistricting battles involved Republican governors fighting their own party. In Florida, Governor Ron DeSantis wanted (and got) a map that elected up to four more Republicans than the legislature’s plan. In New Hampshire, Governor Chris Sununu got a map that made a safe Republican seat more competitive for Democrats. In Wisconsin, Democratic governor Tony Evers didn’t get his way – but the old map, which was Republican-drawn in 2011, was used. In these three states, the net effect of eliminating the governor would be a net gain of up to three more Congressional seats for Democrats.
In summary, almost any interpretation of the Independent State Legislature theory would have effects on Congressional redistricting over the next few years that are close to neutral on average, but often favor Democrats. Such an outcome may give conservatives on the Supreme Court pause.
How is it that such a theory could help Democrats and hurt Republicans? The answer lies in the last decade. In past decades, much reform energy has come from a variety of states, including swing states, where partisan gerrymandering pays off the most. Republicans have taken a more aggressive approach to redistricting, and in many places they have already maxed out their gains.
But neither party should want this change, which replaces a longstanding system of court intervention with a Wild West scenario. This Supreme Court case has profound long-term consequences that go well beyond Congressional districting.
State-level action is the one major route left for reformers to check the runaway power of legislatures, which themselves are often gerrymandered. More broadly, as my colleagues and I have written, every state constitution has provisions that can be used to protect voting rights. Ballot measures like ranked-choice voting can increase competition and de-emphasize extreme candidates. And the theory has potential implications for the selection of Presidential electors. If the Supreme Court blocks these uses of federalism, recovering and strengthening our troubled national democracy will become much harder.
*Arizona, California, Florida, Maryland, Michigan, New Hampshire, New York, North Carolina, Pennsylvania, and Wisconsin.
**The Gerrymandering Project didn’t do simulations in California. With so many districts, the compute time was too long to meet the Project’s goal of doing 1 million simulations in each state. Instead one can use statewide metrics (though these are imperfect by themselves) and example hypothetical maps that show extremes.